mic
6th August 2005, 22:08
Trying to get higher quality audio in captures from digital cable, came up with a method that seems to work well, though it is a bit of extra work -- thought it might provide some ideas for anyone that might be interested .
The ideal would be to record the ac3 stream as broadcast, but not possible with the hardware I've got. As 2nd best I record the ac3 stream decoded by my Audigy2. The advantage is much cleaner audio. Multi-channel capture is possible, but only to the 2 gig file limit of the .wav format, though someone might find a way around this.
The disadvantage I've encountered regards timing, and it can be a big hurdle -- the audio can go out of sync. -- if you're lucky not enough that you'd notice. I use Vegas to adjust the audio track, time stretching or compressing the audio as needed. Normally I'll capture twice, once in the ATI VCR format so I have CC data to work with for subs, & once in whatever capture format for higher quality, so in my situation I have a 2nd audio track to use as a guide in adjusting the decoded ac3 track. [see below *** for more info]
The steps to record decoded ac3 aren't that bad. First off you need to be able to receive the spdif signal, and with my common Motorola digital cable box, that meant a signal that is way off spec.. Fortunately Soundblasters can handle this if you use the on-card digital CD in, which saves from buying/making a convertor box to get a standard signal.
If you have a card like the Audigy 2 without the front panel or breakout box, browse around here http://www.driverheaven.net/forumdisplay.php?f=56 for driver software & info on jumping 2 pins to turn spdif & other features on. It's well worth it in any case.
While I had the Audigy2 out to jumper the 2 pins, I ran a 2-pin cable (the sort that connects to DVD/CD drives) from outside the case to the CD Digital in on the card. It's simple enough to make up a connector or jumper that goes from this 2 pin cable to an RCA jack, or just solder an RCA jack to the cable, and then run a standard RCA cable to the co-ax spdif out jack on the cable box.
To do the actual recording first open up the creative device controls and make sure spdif is set to decode on the Digital Input tab, and Use Installed Decoder is selected on the Decoder tab. Then I use the Creative MediaSource Player to record, though I've had success with other software -- other apps might work better, but I generally use this one for the auto volume setting, plus the options of using EAX &/or CMSS.
For input I use "What U Hear", though you have to be careful if you have any other inputs connected. You can use the Multi-Channel output option, often in combo with cmss, but this gets you a 6 channel, 24/96 file that stops in about 20 minutes or so because of the .wav file limit. And the CMS player should run & record right alongside your capture software as it does it's thing with the video.
Is it worth it? I'd say so -- for me it's a huge improvement. And in my situation it's really not any more work, especially when you consider that recording normally from my digital cable box I had to run 2 passes with the Sony Noise Reduction filter, normalize, run another noise reduction pass, & then run a noise gate pass, and still got much lower quality results.
Now there are quite a few variables, so I present this all as a starting point. Your capture software might record the decoded ac3 right along with the video, but ATI's MMC for example will not on my system. You might go with the KX driver set, which might allow you to capture the raw ac3 file, but it won't go with ATI's MMC -- I use the YouP drivers.
*** Audio & video use different methods of timing, so any time you do separate recordings you run extra risk of losing sync. -- in fact that's why I started using Vegas years ago -- it's the easiest way I know of to stretch or compress the length of an audio track.
Whether matching patterns in .wav files or syncing with the video, I make sure the start of the audio is in sync., then start moving down the timeline to find the first instance where the audio is out of sync by more then a few frames. I find a relatively quiet spot, then split the audio track. I take the portion of the track on the right, and move it back into sync, paying attention to the far left of this trimmed sound clip. Then I take the portion of the track on the left, the part I cut off, and stretch it so both pieces butt together. (In Vegas stretching is accomplished just by ctrl - dragging the end of the clip.)
Progressing through the project, moving from left to right, I just repeat the process if I find things out of sync by more then a few frames.
The ideal would be to record the ac3 stream as broadcast, but not possible with the hardware I've got. As 2nd best I record the ac3 stream decoded by my Audigy2. The advantage is much cleaner audio. Multi-channel capture is possible, but only to the 2 gig file limit of the .wav format, though someone might find a way around this.
The disadvantage I've encountered regards timing, and it can be a big hurdle -- the audio can go out of sync. -- if you're lucky not enough that you'd notice. I use Vegas to adjust the audio track, time stretching or compressing the audio as needed. Normally I'll capture twice, once in the ATI VCR format so I have CC data to work with for subs, & once in whatever capture format for higher quality, so in my situation I have a 2nd audio track to use as a guide in adjusting the decoded ac3 track. [see below *** for more info]
The steps to record decoded ac3 aren't that bad. First off you need to be able to receive the spdif signal, and with my common Motorola digital cable box, that meant a signal that is way off spec.. Fortunately Soundblasters can handle this if you use the on-card digital CD in, which saves from buying/making a convertor box to get a standard signal.
If you have a card like the Audigy 2 without the front panel or breakout box, browse around here http://www.driverheaven.net/forumdisplay.php?f=56 for driver software & info on jumping 2 pins to turn spdif & other features on. It's well worth it in any case.
While I had the Audigy2 out to jumper the 2 pins, I ran a 2-pin cable (the sort that connects to DVD/CD drives) from outside the case to the CD Digital in on the card. It's simple enough to make up a connector or jumper that goes from this 2 pin cable to an RCA jack, or just solder an RCA jack to the cable, and then run a standard RCA cable to the co-ax spdif out jack on the cable box.
To do the actual recording first open up the creative device controls and make sure spdif is set to decode on the Digital Input tab, and Use Installed Decoder is selected on the Decoder tab. Then I use the Creative MediaSource Player to record, though I've had success with other software -- other apps might work better, but I generally use this one for the auto volume setting, plus the options of using EAX &/or CMSS.
For input I use "What U Hear", though you have to be careful if you have any other inputs connected. You can use the Multi-Channel output option, often in combo with cmss, but this gets you a 6 channel, 24/96 file that stops in about 20 minutes or so because of the .wav file limit. And the CMS player should run & record right alongside your capture software as it does it's thing with the video.
Is it worth it? I'd say so -- for me it's a huge improvement. And in my situation it's really not any more work, especially when you consider that recording normally from my digital cable box I had to run 2 passes with the Sony Noise Reduction filter, normalize, run another noise reduction pass, & then run a noise gate pass, and still got much lower quality results.
Now there are quite a few variables, so I present this all as a starting point. Your capture software might record the decoded ac3 right along with the video, but ATI's MMC for example will not on my system. You might go with the KX driver set, which might allow you to capture the raw ac3 file, but it won't go with ATI's MMC -- I use the YouP drivers.
*** Audio & video use different methods of timing, so any time you do separate recordings you run extra risk of losing sync. -- in fact that's why I started using Vegas years ago -- it's the easiest way I know of to stretch or compress the length of an audio track.
Whether matching patterns in .wav files or syncing with the video, I make sure the start of the audio is in sync., then start moving down the timeline to find the first instance where the audio is out of sync by more then a few frames. I find a relatively quiet spot, then split the audio track. I take the portion of the track on the right, and move it back into sync, paying attention to the far left of this trimmed sound clip. Then I take the portion of the track on the left, the part I cut off, and stretch it so both pieces butt together. (In Vegas stretching is accomplished just by ctrl - dragging the end of the clip.)
Progressing through the project, moving from left to right, I just repeat the process if I find things out of sync by more then a few frames.